Discussion:
[ofw] Opensm & WinMad: a race, causing BSOD722
Smith, Stan
2012-01-27 18:11:44 UTC
Permalink
Data point: the head of SVN builds and installs correctly and validates correctly for Connect-X2 and Infinihost on the 6 systems (2 CX + 4 InfiniHost) I tested.
-----Original Message-----
From: Hefty, Sean
Sent: Thursday, January 26, 2012 6:06 PM
To: Tzachi Dar; Leonid Keller; Smith, Stan
Cc: Uri Habusha; ofw_list; Irena Gannon
Subject: RE: Opensm & WinMad: a race, cauing BSOD722
I swapped the HCA with one from another system, but still see the error. The
OFED 3.0 release works fine. I will try backing out my changes and testing
with the head of the svn tree tomorrow or early next week, though I don't see
why the changes would affect loading the driver.
Ok, my build went faster than I expected. I was able to apply this patch to the 3.0 code release and verify that it at least doesn't crash the
system.
- Sean
Hefty, Sean
2012-01-27 18:15:56 UTC
Permalink
Post by Smith, Stan
Data point: the head of SVN builds and installs correctly and validates
correctly for Connect-X2 and Infinihost on the 6 systems (2 CX + 4 InfiniHost)
I tested.
The HCA information from my system is below. Stan, if you can send me a pointer offline, I'll try installing using your build to verify that I wasn't picking up some old installation.

C:\Users\Administrator>ibv_devinfo -v
hca_id: ibv_device0
fw_ver: 0x200070000
node_guid: 0002:c903:000e:c3fe
sys_image_guid: 0002:c903:000e:c401
vendor_id: 0x02c9
vendor_part_id: 26428
hw_ver: 0xB0
phys_port_cnt: 2
max_mr_size: 0xffffffffffffffff
page_size_cap: 0x1000
max_qp: 131000
max_qp_wr: 16351
device_cap_flags: 0x00005876
max_sge: 32
max_sge_rd: 0
max_cq: 65408
max_cqe: 4194303
max_mr: 262128
max_pd: 32764
max_qp_rd_atom: 16
max_ee_rd_atom: 0
max_res_rd_atom: 128
max_qp_init_rd_atom: 128
max_ee_init_rd_atom: 0
atomic_cap: ATOMIC_HCA (1)
max_ee: 0
max_rdd: 0
max_mw: 0
max_raw_ipv6_qp: 0
max_raw_ethy_qp: 0
max_mcast_grp: 8192
max_mcast_qp_attach: 56
max_total_mcast_qp_attach: 458752
max_ah: 0
max_fmr: 0
max_srq: 65472
max_srq_wr: 16383
max_srq_sge: 31
max_pkeys: 128
local_ca_ack_delay: 15
port: 1
state: PORT_ACTIVE (4)
max_mtu: 4096 (5)
active_mtu: 4096 (5)
sm_lid: 1
port_lid: 10
port_lmc: 0x00
max_msg_sz: 0x40000000
port_cap_flags: 0x00005080
max_vl_num: 2 (2)
bad_pkey_cntr: 0x0
qkey_viol_cntr: 0x0
sm_sl: 0
pkey_tbl_len: 16
gid_tbl_len: 128
subnet_timeout: 18
init_type_reply: 0
active_width: 4X (2)
active_speed: 10.0 Gbps (4)
phys_state: LINK_UP (5)
GID[ 0]: fe80:0000:0000:0000:0002:c903:00
0e:c3ff

port: 2
state: PORT_DOWN (1)
max_mtu: 4096 (5)
active_mtu: 4096 (5)
sm_lid: 0
port_lid: 0
port_lmc: 0x00
max_msg_sz: 0x40000000
port_cap_flags: 0x00005080
max_vl_num: 2 (2)
bad_pkey_cntr: 0x0
qkey_viol_cntr: 0x0
sm_sl: 0
pkey_tbl_len: 16
gid_tbl_len: 128
subnet_timeout: 0
init_type_reply: 0
active_width: 4X (2)
active_speed: 2.5 Gbps (1)
phys_state: POLLING (2)
GID[ 0]: fe80:0000:0000:0000:0002:c903:00
0e:c400
Hefty, Sean
2012-01-27 22:30:34 UTC
Permalink
I updated to FW 2.9.1 and the tip of svn worked again. IMO, the latest drivers should not automatically break a working HCA just because the firmware is older (was 2.7).
-----Original Message-----
From: Smith, Stan
Sent: Friday, January 27, 2012 10:12 AM
To: Hefty, Sean; Tzachi Dar; Leonid Keller
Cc: Uri Habusha; ofw_list; Irena Gannon
Subject: RE: Opensm & WinMad: a race, causing BSOD722
Data point: the head of SVN builds and installs correctly and validates
correctly for Connect-X2 and Infinihost on the 6 systems (2 CX + 4 InfiniHost)
I tested.
-----Original Message-----
From: Hefty, Sean
Sent: Thursday, January 26, 2012 6:06 PM
To: Tzachi Dar; Leonid Keller; Smith, Stan
Cc: Uri Habusha; ofw_list; Irena Gannon
Subject: RE: Opensm & WinMad: a race, cauing BSOD722
I swapped the HCA with one from another system, but still see the error.
The
OFED 3.0 release works fine. I will try backing out my changes and testing
with the head of the svn tree tomorrow or early next week, though I don't
see
why the changes would affect loading the driver.
Ok, my build went faster than I expected. I was able to apply this patch to
the 3.0 code release and verify that it at least doesn't crash the
system.
- Sean
Loading...